The role of polysaccharides and diatom exudates

The role of polysaccharides and diatom exudates

1

in the redox cycling of Fe and the photoproduction of

2

hydrogen peroxide in coastal seawaters

3

Sebastian Steigenberger¹, Peter J. Statham², Christoph Völker¹ and Uta Passow¹ 4

5

1Alfred Wegener Institut für Polar- und Meeresforschung, Am Handelshafen 12, 6

27570 Bremerhaven, Germany 7

2National Oceanography Centre, Southampton, University of Southampton Waterfront 8

Campus, European Way, Southampton SO14 3ZH 9

10

Abstract 11

The effect of artificial acidic polysaccharides (PS) and exudates of 12

Phaeodactylum tricornutum on the half-life of Fe(II) in seawater was investigated in 13

laboratory experiments. Strong photochemical hydrogen peroxide (H2O2) production 14

of 5.2 to 10.9 nM (mg C)^-1 h^-1 was found in the presence of PS and diatom exudates.

15

Furthermore when illuminated with UV light algal exudates kept the concentration of 16

ferrous iron in seawater (initial value 100 nmol L^-1) elevated for about 50 min. Since 17

no stabilising effect of PS on Fe(II) in the dark could be detected, enhanced 18

photoreduction seems to be the cause. This was confirmed by a simple model of the 19

photochemical redox cycle of iron. Diatom exudates seem to play an important role 20

for the photochemistry of iron in coastal waters.

21

22

1 Introduction 23

Marine phytoplankton contributes significantly to the CO2 exchange between 24

atmosphere and ocean, thus impacting atmospheric CO2 concentrations (Falkowski et 25

al. 1998). Global marine primary productivity shows great spatial and temporal 26

variability, caused primarily by variable light, zooplankton grazing and nutrient 27

distributions. In addition to the macronutrients (P, N), iron is an essential trace 28

element for photo-autotrophic organisms (Geider et al. 1994; Falkowski et al. 1998;

29

Morel et al. 2003). Several large scale iron fertilization experiments have revealed 30

that in 40% of the surface ocean, the so called High Nutrient Low Chlorophyll 31

(HNLC) areas, iron is at least partially responsible for limitation of phytoplankton 32

growth (Boyd et al. 2007). However, iron limitation can occur in coastal areas as well 33

(Hutchins et al. 1998) and here the supply of Fe through upwelling and resuspension 34

determine its cycling.

35

Free hydrated Fe(III) concentrations in seawater are very low (<10^-20 mol L^-1) (Rue et 36

al. 1995) and the more soluble Fe(II) is rapidly oxidised (Millero et al. 1987; Millero 37

et al. 1989; King et al. 1995; Gonzalez-Davila et al. 2005, 2006). Thus concentrations 38

of dissolved Fe in the ocean should be very low. However, over 99% of the dissolved 39

iron in seawater is reported to be bound by organic compounds (Rue et al. 1995; van 40

den Berg 1995; Croot et al. 2000; Boye 2001) and these ligands can maintain the 41

concentrations typically seen in the ocean (Johnson et al. 1997). Iron binding ligands 42

in seawater mainly consist of bacterial siderophores (Macrellis et al. 2001; Butler 43

2005) and possibly planktonic exudates like acidic polysaccharides (PS) (Tanaka et 44

al. 1971). Transparent exopolymer particles (TEP), which are rich in acidic 45

polysaccharides, are ubiquitous in the surface ocean (Passow 2002). TEP has been 46

shown to bind ²³⁴Th (Passow et al. 2006) and are therefore a prime candidate to bind 47

iron.

48

The main oxidation pathway of Fe(II) to Fe(III) is the reaction with O2 and 49

H2O2 according to the Haber-Weiss mechanism (Millero et al. 1987; Millero et al.

50

1989; King et al. 1995). This oxidation can be inhibited (Theis et al. 1974; Miles et 51

al. 1981) or accelerated (Sedlak et al. 1993; Rose et al. 2002, 2003a) in the presence 52

of organic compounds. The decrease in apparent oxidation rate is suggested to be due 53

to stronger photoreduction of Fe(III) (Kuma et al. 1995) or stabilisation of Fe(II) 54

(Santana-Casiano et al. 2000; Rose et al. 2003b; Santana-Casiano et al. 2004).

55

In marine systems H2O2 functions as a strong oxidant or a reductant (Millero 56

et al. 1989; Croot et al. 2005). Thus it is important for the cycling of organic 57

compounds and trace metals like Fe (Millero et al. 1989). H2O2 is the most stable 58

intermediate in the reduction of O2 to H2O and is mainly produced in the water 59

column by photochemical reactions involving dissolved organic matter (DOM) and 60

O2 (Cooper et al. 1988; Scully et al. 1996; Yocis et al. 2000; Yuan et al. 2001). Light 61

absorbed by DOM induces an electron transfer to molecular oxygen, forming the 62

superoxide anion radical, which undergoes disproportionation to form hydrogen 63

peroxide. Hence light, O2, H2O2 and organic compounds are important factors in the 64

very complex chemistry of iron in seawater.

65

Increased photochemical reduction of Fe(III) in the presence of sugar acids has 66

been reported (Kuma et al. 1992; Ozturk et al. 2004; Rijkenberg et al. 2005) but for 67

polysaccharides no such studies have been carried out so far. However, the relative 68

abundance of polysaccharides in marine dissolved organic matter (DOM) is about 69

50% (Benner et al. 1992) and in phytoplankton derived DOM the fraction of 70

polysaccharides can be up to 64% (Hellebust 1965; Hellebust 1974). In the study 71

reported here we investigate the effect of PS and algal exudates on the photochemical 72

redox cycle of iron and production of H2O2. 73

74

2 Materials and Methods 75

2.1 General 76

Three different types of experiments were conducted to investigate the effect 77

of PS and diatom exudates in combination with UV light on the speciation of iron and 78

the production of H2O2. All experiments were conducted at a constant temperature 79

(about 20°C) in the laboratory. In experiments 1 and 3 were samples were exposed to 80

UV radiation, UV transparent 3 L Tedlar bags were used as incubation containers.

81

Experiment 2 was conducted in 30 mL polystyrene screw cap tubes, without UV 82

irradiation.

83

The natural coastal seawater (SW) was collected in July 2006 off Lepe near 84

Southampton (UK), filtered through 0.2 µm membranes and stored at 5°C. Organic 85

matter was removed from a part of this SW via photo-oxidation with strong UV 86

radiation. The so called “organic-free” UVSW (Donat et al. 1988) was also stored at 87

5°C.

88

We used gum xanthan, laminarin and carrageenan (all from Sigma) as the 89

artificial PSs. The molecular weight of laminarin is 7700 g mol^-1 (Rice et al. 2004) 90

and 43% (w/w) of the molecule is carbon. For gum xanthan and carrageenan no 91

specifications could be found but we assumed a carbon content of ~40% (w/w).

92

Diatom exudates were collected as the 0.4 µm filtrate of a senescent culture of 93

Phaeodactylum tricornutum grown in f/2 medium. Ford and Percival (1965) separated 94

a significant amount of a water-soluble glucan from an aqueous extract of 95

Phaeodactylum tricornutum, and their results showed this polysaccharide to be a 96

typical chrysolaminarin with essential similar properties to the p-1,3-linked glucan, 97

laminarin.

98

Philips 40TL12 and Philips 40T’05 lamps, respectively, were used as a light 99

source for the irradiation of samples with UVB and UVA light during experiments 1 100

and 3. Irradiance was measured with a UVA (315-400 nm) sensor type 2.5, a UVB 101

(280-315 nm) sensor type 1.5 (INDIUM-SENSOR, Germany) and a spherical 102

quantum sensor SPQA 2651 (LI-COR) for the photosynthetically active radiation 103

(PAR, 400-700 nm). Sensors were coupled to a data logger LI-1400 (LI-COR). The 104

following irradiance values were used for all light incubations during this study:

105

UVB=0.3 W m^-2, UVA=17.6 W m^-2 and PAR=3.8 W m^-2. For all experiments 106

samples were held in UV transparent 3 L polyvinyl fluoride (PVF, Tedlar) bags (SKC 107

Inc., USA), fitted with a polypropylene hose for filling and sub-sampling.

108

109

2.2 Specific Experiments 110

2.2.1 Experiment 1: Effect of polysaccharides on the photogeneration of H2O2

111

Four pairs of Tedlar bags were filled with MQ water and concentrated 112

solutions of three different PSs were added to three pairs of these bags. For this 113

experiment carrageenan, gum xanthan and laminarin were used. The PSs were 114

dissolved in MQ water by sonicating for 30 min. The final concentration of PS was 115

10 mg L^-1 in about 2.3 L. The last pair of bags served as control and contained no PS.

116

One bag of each pair was placed in the dark the other was illuminated with UV light 117

for 270 min. H2O2 was measured 1 h before illumination and after 0, 10, 30, 90, 118

270 min in the light and the dark sample.

119

120

2.2.2 Experiment 2: Effect of polysaccharides on the oxidation of Fe(II) in seawater 121

in the dark 122

Ten clean polystyrene screw cap tubes (30 mL) were filled with the natural 123

Solent seawater (0.2 µm filtered) and another ten tubes were filled with the organic- 124

free Solent Seawater. To 5 tubes of each treatment gum xanthan was added to a final 125

concentration of 1 mg L^-1 and the samples were sonicated for 30 min. Initially Fe(II) 126

equivalent to 200 nmol L^-1 was added to all tubes, and Fe(II) and H2O2 measured after 127

0, 2, 6, 18, 54 min. Temperature, salinity, oxygen concentration and pH were 128

measured before the iron addition and at the end of the experiment.

129

130

2.2.3 Experiment 3: Effect of diatom exudates and UVA/B radiation on the oxidation 131

of Fe(II) in seawater 132

Three Tedlar bags were filled with about 1 L of organic-free seawater (0.2 µm 133

filtered). One bag served as a control and no further additions were made. To the 134

second bag 100 nmol L^-1 Fe(II) were added. To the third bag an addition of diatom 135

exudates and 100 nmol L^-1 Fe(II) was made. The amount of diatom exudates added to 136

the sample was chosen in order to reach a concentration of PS similar to natural 137

over a 60 min period after the iron addition. The UV light was switched on for the 139

whole experiment right after the addition of iron to the sample bags. Temperature, 140

salinity, oxygen concentration, pH and total iron were measured before the iron 141

addition and at the end of the experiment. H2O2 in the organic-free seawater was 142

adjusted to an initial concentration of 5 nmol L^-1 and was measured again at the end of 143

the experiment.

144

145

2.3 Analyses 146

Iron concentrations in the samples were determined using a colorimetric 147

method described by Stookey (1970) and Viollier et al. (2000). Briefly Ferrozine (the 148

disodium salt of 3-(2-pyridyl)-5,6-bis(4-phenylsulfonic acid)-1,2,4-triazine) forms a 149

magenta coloured tris complex with ferrous iron. The water soluble complex is stable 150

and quantitatively formed in a few minutes at pH = 4-9 after adding an aqueous 151

0.01 mol L^-1 Ferrozine solution. The absorbance was measured with a Hitachi U-1500 152

at 562 nm in 10 cm cuvettes buffered with an ammonium acetate buffer adjusted to 153

pH = 5.5, and compared to a calibration curve made by standard additions to the 154

sample water. Standards were prepared from a 10 mmol L^-1 Fe(II) stock solution 155

(Fe(NH4)2(SO4)2.6H2O in 0.1 mol L^-1 HCl) diluted in 0.01 mol L^-1 HCl. Total iron 156

was determined by previous reduction of the iron present in the sample under acid 157

conditions over 2 h at room temperature by adding hydroxylamine hydrochloride 158

(1.4 mol L^-1 in 5 mol L^-1 HCl) as the reducing agent. The detection limit of this 159

method is about 8 nmol L^-1 of Fe(II) and the standard error is about 20%. All 160

Reagents were from Sigma-Aldrich and at least p.a. grade. All solutions were 161

prepared in MQ water (18 MΩ cm^-1) purified with a Millipore deionisation system.

162

Samples were prepared in 30 mL polystyrene screw cap tubes. All equipment has 163

been carefully acid washed prior to use.

164

Concentrations of dissolved mono- and polysaccharides were determined semi 165

quantitatively using another colorimetric method described by Myklestad et al.

166

(1997). Briefly the absorbance of the strong coloured complex of 2,4,6-tripyridyl-s- 167

triazine (TPTZ) formed with iron reduced by monosaccharides or previously 168

hydrolyzed polysaccharides at alkaline pH is measured at 595 nm in 2.5 cm cuvettes 169

and compared to a calibration curve prepared from D-glucose in MQ water. Total 170

sugar concentration was determined after hydrolysis of the acidified sample in a 171

sealed glass ampoule at 150°C for 90 min. The detection limit was 172

0.02 mg glucose eq. L^-1 and the standard error was about 3%. All glassware and 173

reagents were prepared as described by Myklestad et al. (1997).

174

For the determination of hydrogen peroxide (H2O2) a chemiluminescence flow 175

injection analysis (FIA-CL) described by Yuan and Shiller (1999) was used. The 176

method is based on oxidation of luminol by hydrogen peroxide in an alkaline solution 177

using Co(II) as a catalyst. Our flow injection system generally resembled that 178

described by Yuan and Shiller (1999) but as a detection unit we used the photosensor 179

module H8443 (Hamamatsu) with a power supply and a signal amplifier. The voltage 180

signal was logged every second using an A/D converter and logging software (PMD- 181

1208LS, Tracer DAQ 1.6.1.0, Measurement Computing Corporation). The 182

chemiluminescence peaks were evaluated by calculating their area. The detection 183

limit was 0.1 nmol L^-1 and the standard error was 4%. All reagents and solutions were 184

prepared as described by Yuan and Shiller (1999). Since ferrous iron in the sample 185

shows a significant positive interference (Yuan et al. 1999) H2O2 was measured in 186

parallel samples without added Fe(II) or after one hour when most of the iron was 187

reoxidised.

188

A WTW 315i T/S system was used to determine temperature and salinity in 189

the sample. Oxygen was measured using a WPA OX20 oxygen meter. The dissolved 190

organic carbon (DOC) content in the 0.2 µm filtered samples was measured with a 191

Shimadzu TOC-VCSN system via high temperature catalytic oxidation (HTCO) on Pt 192

covered Al2O3 beads. The detection limit of this method is ~3 µmol L^-1 and the 193

precision is ±2 µmol L^-1. 194

The UV photooxidation system consisted of a fan cooled 1 kW medium 195

pressure mercury lamp (Hanovia), with 10 x 200 mL quartz tubes mounted around the 196

axial lamp. After 6 h of UV irradiation the samples were considered “organic-free”

197

(UVSW) (Donat et al. 1988). To remove the resulting high concentrations of H2O2 the 198

organic-free water was treated with activated charcoal. The charcoal had previously 199

been washed several times with HCl, ethanol and MQ water to remove contaminants.

200

After stirring for 30-40 min the charcoal was removed by filtration through a 0.2 µm 201

polycarbonate membrane. The H2O2 concentration in the resulting water was less than 202

0.5 nmol L^-1 and no contamination with iron was detectable.

203

204

3 Results and discussion 205

3.1 Experiment 1: Effect of polysaccharides on the photochemical production of 206

H2O2

207

The first experiment, examining the effect of polysaccharides on the 208

photochemical production of H2O2, showed that within 270 min (4.5 h) of 209

illumination large amounts (140-240 nmol L^-1) of H2O2 were formed due to the 210

addition of 10 mg L^-1 of polysaccharides to MQ water (Figure 1). The H2O2

211

concentrations in all samples increased linearly during the experiment, after the light 212

was switched on. Gum xanthan showed the highest photochemical production of H2O2

213

followed by carrageenan and laminarin, which can be explained by their different 214

absorptivity at <400 nm (Figure 2). The addition of laminarin led to a net 215

accumulation rate of H2O2 of 22.5 nmol L^-1 h^-1, which was twice as high as that for 216

pure MQ water (12.3 nmol L^-1 h^-1). The H2O2 accumulation during illumination of the 217

MQ water was probably due to organic matter leaching from the resin of the filter 218

cartridge of the MQ system. However, the DOC concentration in MQ water was 219

<<10 µmol L^-1. H2O2 accumulation rates of 36.2 nmol L^-1 h^-1 and 43.4 nmol L^-1 h^-1 220

were determined in samples with added carrageenan and gum xanthan, respectively.

221

The photochemical production of H2O2 was thus 3-4 times higher in the presence of 222

carrageenan and gum xanthan compared to pure MQ water. Linear H2O2

223

accumulation rates of similar magnitude have been reported by Cooper et al. (1988) 224

and Miller et al. (1995) in natural seawater samples. The main structural differences 225

between the molecules of these three PSs are that laminarin has a linear structure of 226

linked glucose monosaccharide units, carrageenan has sulphur containing groups and 227

gum xanthan has a branched structure incorporating uronic acid groups. The PS 228

concentration used in our experiment is equivalent to about 4 mg L^-1 organic carbon 229

leading to normalised H2O2 generation rates of 5.2 nmol L^-1 (mg C)^-1 h^-1 (laminarin), 230

9.1 nmol L^-1 (mg C)^-1 h^-1 (carrageenan) and 10.9 nmol L^-1 (mg C)^-1 h^-1 (gum xanthan).

231

These values are up to 29 times higher than the rate of 0.38 nmol L^-1 (mg C)^-1 h^-1 232

reported by Price et al. (1998) for the >8000 Da fraction of natural DOM in the 233

Western Mediterranean even though the light bulbs used in our study typically 234

produced only 25% of the UVB radiation 39% of UVA and 1% of PAR of the 235

calculated natural irradiance found in midday summer sun in the Mediterranean (Zepp 236

et al. 1977). The polysaccharides in our study caused strong photogeneration of H2O2

237

even under low light exposure probably due to the absence of removal processes such 238

as enzymatic decomposition of H2O2 (Moffett et al. 1990). Photochemical production 239

rates of H2O2 in the Atlantic Ocean and Antarctic waters are much lower ranging from 240

2.1 to 9.6 nmol L^-1 h^-1 (Obernosterer 2000; Yocis et al. 2000; Yuan et al. 2001;

241

Gerringa et al. 2004). Gerringa et al. (2004) calculated a net production rate of 242

7 nmol L^-1 h^-1 at irradiance levels of 2.8 (UVB), 43 (UVA) and 346 W m^-2 (VIS/PAR) 243

in 0.2 µm filtered water from the eastern Atlantic close to the Equator. These low 244

rates are presumably due to lower DOC concentrations and higher decay rates due to 245

colloids or enzymatic activity in natural waters (Moffett et al. 1990; Petasne et al.

246

1997). Our experiments suggest that PSs may have had a significant indirect effect on 247

Fe oxidation due to the enhanced photochemical production of H2O2. 248

249

3.2 Experiment 2: Effect of gum xanthan on the oxidation of Fe(II) in the dark 250

Differences in the rate of Fe(II) oxidation due to added gum Xanthan were 251

small, both in the natural SW and the UVSW samples (Figure 3 and 4). However, the 252

oxidation of Fe(II) in the natural SW samples (with or without gum xanthan) (Figure 253

3) was much slower than that in the respective DOM-free UVSW samples (Figure 4).

254

Half-life values and oxidation rates of organic-free seawater can be calculated 255

according to Millero and Sotolongo (1989) and Millero et al. (1987). Under our 256

experimental conditions the calculated half-life was 25 s for the ambient H2O2

257

concentrations and 82 s under O2 saturation. These theoretical values can be compared 258

to measured Fe(II) half-life values of 42 s (UVSW) and 35 s (UVSW+PS). The 259

measured values resemble the theoretical values under the ambient H2O2 conditions.

260

This indicates that the high H2O2 concentration had a stronger oxidising effect on 261

Fe(II) than the dissolved O2 in the samples.

262

For the natural SW sample the theoretical half-life of 43 s under O2 saturation 263

does not fit the measured data well. The half-life of Fe(II) in the natural SW sample 264

(Figure 3) was ~17 times (11.9 min) and with PS added ~19 times (13.3 min) longer 265

than theoretical value. The measured data followed the exponential oxidation curve 266

calculated for the low H2O2 concentration of these samples whereas the high O2

267

content seemed to not accelerate the measured oxidation of Fe(II).

268

The DOC content of the natural SW (97 µmol L^-1) was almost 10 times higher 269

than of the UVSW. The difference in Fe(II) oxidation between the water types might 270

therefore be due to the stabilisation of Fe(II) against oxidation by natural occurring 271

compounds of the coastal SW (Theis et al. 1974; Miles et al. 1981; Santana-Casiano 272

et al. 2000; Rose et al. 2003a; Santana-Casiano et al. 2004). These results show that 273

the added gum xanthan was not a good model for natural occurring substances 274

stabilising Fe(II) against oxidation. Initial H2O2 concentrations also differed 275

appreciably, with 5 nmol L^-1 H2O2 in the natural SW sample and 270 nmol L^-1 H2O2 in 276

the UVSW sample. UV oxidation in UVSW water during removal of natural DOM 277

must have caused the differences in H2O2. We calculated Fe(II) oxidation rates due to 278

O2 and H2O2 in our experiment to investigate if the differing rates could have been 279

caused by differing initial H2O2 concentrations. From the comparison between our 280

measured and theoretically calculated values we conclude that a strong effect of H2O2

281

on the lifetime of Fe(II) was observed but no effect of gum xanthan was found in this 282

experiment conducted without irradiation. The lower initial H2O2 concentrations in 283

the natural SW sample (5 nmol L^-1 H2O2; Figure 3) compared to the UVSW sample 284

(270 nmol L^-1 H2O2; Figure 4) appears to be the major cause for slower Fe(II) 285

oxidation, suggesting that H2O2 mainly control the oxidation of Fe(II).

286

287

3.3 Experiment 3: Effect of diatom exudates and UVA/B radiation on the oxidation 288

of Fe(II) in seawater 289

Initially, the half-lives of Fe(II) in both treatments, those with and without 290

addition of diatom exudates, was quite similar (Figure 5). For the initial 5 min (300 s) 291

a half life of 4.5±0.7 min and 4.0±0.3 min, respectively was determined for Fe(II) in 292

the UVSW without and with added diatom exudates. These values are in the same 293

range as published values (Millero et al. 1987; Kuma et al. 1995; Croot et al. 2002).

294

A remarkable difference between both treatments is clearly visible after about 7 min 295

(420 s) (Figure 5). In the UVSW without exudates the Fe(II) concentration continued 296

decreasing exponentially reaching the detection limit after 20 min, whereas in UVSW 297

with added diatom exudates the Fe(II) concentration remained at about 30 nmol L^-1 298

decreasing only very slightly with time. The photochemical effect of the exudates was 299

strong enough to result in a net stabilising effect on Fe(II) after 7 minutes.

300

Differences in H2O2 production during the first hour of irradiation were 301

significant between UVSW with and without exudates. In the UVSW sample with 302

added diatom exudates a linear production rate of 33 nmol L^-1 h^-1 H2O2 was 303

determined whereas in pure UVSW the respective rate was only 5 nmol L^-1 h^-1. The 304

higher production rate of H2O2 in the presence of exudates, suggests increased 305

photochemical production of H2O2. UVSW without exudates contained 11 µmol L^-1 306

DOC and no measurable total MS and PS, whereas UVSW mixed with exudates of 307

Phaeodactylum tricornutum contained ~450 µmol L^-1 DOC, including 308

0.4 mg glucose eq. L^-1 (i.e. 13 µmol C L^-1) total MS and PS. The DOC- normalised 309

H2O2 generation rate of 6.1 nmol L^-1 (mg C)^-1 h^-1 calculated from UVSW with 310

exudates indicates that laminarin-like diatom exudates (Ford et al. 1965) 311

photochemically produce H2O2. However, the high DOC content suggests that there 312

was also other organic matter contributing to the photo-production of H2O2. 313

Figure 6 shows a schematic of that part of the iron cycle relevant for our 314

experiment. In pure UVSW the added Fe(II) was oxidised rapidly, but in the presence 315

of ligands contained in the diatom exudates Fe(II) formed FeL, which in the light was 316

released as Fe(II) and then oxidised. The Fe(II) concentration could thus remain stable 317

as Fe(II) production from FeL balanced Fe(II) oxidation. We used a simple numerical 318

model based on these processes to model the Fe(II) concentration in our experimental 319

system.

320

The model uses a constant photoproduction term khν[FeL] of ferrous iron, and 321

constant oxidation rates with oxygen (kO2). The oxidation rates with hydrogen 322

peroxide (kH2O2) are assumed to increase linearly with a photoformation rate of 323

33 nmol L^-1 h^-1 as measured in this experiment and initial H2O2 concentration are set 324

at 4.6 nmol L^-1. The initial Fe(II) concentration [Fe(II)0] is set at 100 nmol L^-1 Fe(II), 325

the amount added in the experiment, and increases in the model by the constant 326

photoreduction of the FeL complex (where L is either EDTA or diatom exudates or a 327

combination of both). The direct photoreduction of inorganic iron colloids and 328

dissolved ferric iron is also possible (Waite et al. 1984; Wells et al. 1991a; Wells et 329

al. 1991b; Johnson et al. 1994), but rates for these processes are negligibly low. For 330

both processes together we calculated about 0.004 nmol L^-1 s^-1 of Fe(II) for 331

100 nmol L^-1 Fe(II) added using the rates reported by Johnson et al. (1994). The 332

model assumes that the concentration of FeL changes only negligibly during the 333

experiment. As loss processes of Fe(II) we included the oxidation of Fe(II) with O2

334

and the oxidation with H2O2. The latter depends on the increasing H2O2

335

concentrations during the experiment. Since dissociation and formation of FeL are 336

relatively slow (Hudson et al. 1992) compared to the photoreduction of FeL and the 337

oxidation of Fe(II) we ignored these processes in the model. The model calculates the 338

change in Fe(II) concentration over time (equation 1).

339

[

] [ ] [

] [

][

]

2 2

2 Fe II k H O Fe II

k FeL dt k

II Fe d

O H O

hv − −

= eq. 1

340

[

]

341

t given in [s], khν and k^O2 in [s^-1], k^H2O2in [L nmol^-1 s^-1] and all concentrations given in 342

[nmol L^-1].

343

The parameters kO2, khν [FeL] and kH2O2were estimated by fitting the model to the 344

observed data, minimizing the root mean squared model-data misfit, scaled by the 345

assumed variance of the measurements. If the deviations between model and data are 346

independent and normally distributed, the misfit 347

∑

=

i i

i

i m

d

2 2

2 ( )

χ σ eq. 3

348

is a χ² variable. In this case we can estimate the posterior probability density function 349

(pdf) of the model parameters (assuming a uniform prior) by 350

( [ ] )

⎠

⎜⎜ ⎞

⎝

⎛ − exp 2

~ ,

,

2

2 2 2

χ

ν HO

h

O k FeL k

k

pdf eq. 4

351

(see e.g. D.S. Sivia (2006)). The probability function is well approximated by a 352

multidimensional Gaussian distribution with a maximum value for the best estimated 353

set of parameter values. To obtain an estimate of the variance for this maximum 354

likelihood estimate of the parameters, we also need an estimate of the covariance 355

matrix of the parameters at the minimum of χ². This covariance matrix can be 356

estimated as the inverse of the Hessian matrix of χ² at the minimum. We can then 357

assume a confidence interval (± one standard deviation) for the best estimates of the 358

parameters, which are kO2 = 6.04e-03±1.20e-03 s^-1, kH2O2 = 1.97e-04±8.59e-05 359

L nmol^-1 s^-1and khν [FeL] = 0.22±0.06 nmol L^-1 s^-1. With this high photoreduction rate 360

the model fits the measured data very well (Figure 7) but the oxidation rates for 361

oxygen and H2O2 are 30% lower and 105% higher, respectively, than rates reported 362

by Millero et al. (1987; 1989). Holding the oxidation rates kO2 and kH2O2 fixed at 363

values calculated for the given experimental conditions (22 °C, S = 34.2, O2 saturated, 364

pH = 8.1) according to Millero et al. (1987; 1989) the model-data misfit becomes 365

somewhat larger and the model requires a slightly higher Fe(II) photoproduction term 366

khν [FeL] of about 0.24±0.01 nmol L^-1 s^-1 to fit the measured data (Figure 7). The 367

larger error margins when fitting all three parameters, compared to fitting only the 368

photoreduction rate, is explained by the strong correlation between the estimates of 369

kH2O2 and of khν [FeL], meaning that the data can be represented almost equally well 370

with different combinations of these two parameters.

371

The estimated photoproduction rates of Fe(II) are about 50 times higher than the 372

photoreduction rate of inorganic colloidal and dissolved iron calculated before, 373

independent of whether we assume the oxidation rates by Millero et al. (1987, 1989).

374

This indicates high photoreduction of Fe(III) mediated by the added organic material.

375

This high reduction of Fe(III) could have resulted either from direct photoreduction of 376

the FeL or indirectly via light induced (see absorbance spectra Figure 2) formation of 377

superoxide (DOM + hν → DOM*; DOM* + O2 → DOM⁺ + O2¯; and Fe(III) + O2¯ 378

→ Fe(II) + O2) and the subsequent reduction of ferric iron (King et al. 1995; Voelker 379

et al. 1995; Rose et al. 2005; Fujii et al. 2006; Rose et al. 2006; Waite et al. 2006;

380

Garg et al. 2007b, 2007a).

381

Since the estimated laminarin concentration of ~1 mg L^-1 only accounts for 382

~8% of the DOC content of this sample it is not clear to what extend PS were 383

responsible for the photoreduction during this experiment. Some EDTA 384

(concentration of ~1 µmol L^-1) had inadvertently also been added with the diatom 385

exudates, as it was part of the culture media. However, photoreduction of iron from 386

complexes with EDTA seemed to have had only a minor effect. Reported steady state 387

Fe(II) concentrations present under stronger irradiation due to photoreduction of Fe- 388

EDTA complexes are much lower (Sunda et al. 2003) than observed in this study.

389

Photo-redox cycling of Fe–EDTA complexes has a larger influence on Fe(III) 390

concentrations than on those of Fe(II) (Sunda et al. 2003).

391

Steady state concentrations of photochemical Fe(II) are linearly related to the 392

irradiation energy especially in the UV range (Kuma et al. 1995; Rijkenberg et al.

393

2005; Rijkenberg et al. 2006; Laglera et al. 2007). In our study the light intensity was 394

only 25% of the UVB radiation 39% of UVA and 1% of PAR of the calculated natural 395

irradiance in midday summer sun at 40°N (Zepp et al. 1977). Therefore under natural 396

coastal conditions, with 4-5 times lower DOC concentrations but a 2.6 to 100 times 397

higher irradiance levels, a photoreductive effect of diatom exudates seems highly 398

probable.

399

400

4 Conclusions 401

In this study we investigated the photochemical effect of artificial and natural 402

polysaccharide material in aquatic systems on iron speciation and on the production of 403

H2O2. Artificial PS caused high photochemical production of H2O2, which acts as a 404

strong oxidant for metals and organic matter on the one hand. On the other hand H2O2

405

is formed photochemically via the superoxide intermediate which is capable of 406

reducing Fe(III). We found increased steady state Fe(II) concentrations in illuminated 407

seawater with a high concentration of exudates of Phaeodactylum tricornutum. In the 408

dark this effect of artificial PS on ferrous iron was not detectable, suggesting that 409

light-produced superoxide reduces Fe(III) maintaining elevated Fe(II) concentration.

410

In coastal seawater with high content of organic matter originating partly from 411

diatoms a positive effect of the exudates on the bioavailability of iron seems likely.

412

Field studies comparing natural phytoplankton bloom waters with open ocean waters 413

are needed to confirm these photoreduction results and the counteracting effect of 414

H2O2 on a daily time scale and as a function of particle size (dissolved, colloidal and 415

particulate fraction).

416

417

5 Acknowledgments 418

We thank P. Gooddy for his help in the laboratory at the NOCS (UK) 419

and T. Steinhoff and S. Grobe who measured the DOC in our samples at the IfM- 420

Geomar (Germany). Thanks also to N. McArdle for administrational help during this 421

BIOTRACS Early-Stage Training (EST) Fellowship which was funded by the 422

European Union under the Sixth Framework Marie Curie Actions.

423

424

6 References 425

Benner, R., J. D. Pakulski, et al. (1992). "Bulk chemical characteristics of dissolved 426

organic matter in the ocean." Science 255: 1561-1564.

427 428

Boyd, P. W., T. Jickells, et al. (2007). "Mesoscale iron enrichment experiments 1993–

429

2005: synthesis and future directions." Science 315: 612-617.

430 431

Boye, M. (2001). "Organic complexation of iron in the Southern Ocean." Deep Sea 432

Research I 48(6): 1477-1497.

433 434

Butler, A. (2005). "Marine Siderophores and Microbial Iron Mobilization." BioMetals 435

18(4): 369-374.

436 437

Cooper, W. J., R. G. Zika, et al. (1988). "Photochemical formation of H2O2 in natural 438

waters exposed to sunlight." Environ. Sci. Technol. 22: 1156-1160.

439 440

Croot, P. L. and M. Johansson (2000). "Determination of iron speciation by cathodic 441

stripping voltammetry in seawater using the competing ligand 2-(2-Thiazolylazo)-p- 442

cresol (TAC)." Electroanalysis 12(8): 565-576.

443 444

Croot, P. L. and P. Laan (2002). "Continuous shipboard determination of Fe(II) in 445

polar waters using flow injection analysis with chemiluminescence detection."

446

Analytica Chimica Acta 466: 261-273.

447 448

Croot, P. L., P. Laan, et al. (2005). "Spatial and temporal distribution of Fe(II) and 449

H2O2 during EisenEx, an open ocean mescoscale iron enrichment." Mar. Chem. 95:

450

65-88.

451 452

Donat, J. R. and K. W. Bruland (1988). "Direct determination of dissolved Cobalt and 453

Nickel in seawater by differential pulse cathodic stripping voltammetry preceded by 454

adsorptive collection of cyclohexane-1,2-dione dioxime complexes." Anal. Chem. 60:

455

240-244.

456 457

Falkowski, P. G., R. T. Barber, et al. (1998). "Biogeochemical controls and feedbacks 458

on ocean primary production." Science 281(5374): 200-206.

459 460

Ford, C. W. and E. Percival (1965). "The carbohydrates of Phaeodactylum 461

tricornutum." J. Chem. Soc.: 7035-7041.

462 463

Fujii, M., A. L. Rose, et al. (2006). "Superoxide-mediated dissolution of amorphous 464

ferric oxyhydroxide in seawater." Environmental Science & Technology 40(3): 880- 465

887.

466 467

Garg, S., A. L. Rose, et al. (2007a). "Superoxide-mediated reduction of organically 468

complexed iron(III): Impact of pH and competing cations (Ca2+)." Geochimica Et 469

Cosmochimica Acta 71: 5620-5634.

470

471

Garg, S., A. L. Rose, et al. (2007b). "Superoxide mediated reduction of organically 472

complexed Iron(III): Comparison of non-dissociative and dissociative reduction 473

pathways." Environmental Science & Technology 41(9): 3205-3212.

474 475

Geider, R.J, et al. (1994). "The role of iron in phytoplankton photosynthesis and the 476

potential for iron-limitation of primary productivity in the sea." Photosynth. Res. 39:

477

275-301.

478 479

Gerringa, L. J. A., M. J. A. Rijkenberg, et al. (2004). "The influence of solar 480

ultraviolet radiation on the photochemical production of H2O2 in the equatorial 481

Atlantic Ocean." J. Sea Res. 51: 3-10.

482 483

Gonzalez-Davila, M., J. M. Santana-Casiano, et al. (2005). "Oxidation of iron (II) 484

nanomolar with H2O2 in seawater." Geochimica et Cosmochimica Acta 69(1): 83-93.

485 486

Gonzalez-Davila, M., J. M. Santana-Casiano, et al. (2006). "Competition between O2

487

and H2O2 in the oxidation of Fe(II) in natural waters." Journal of Solution Chemistry 488

35(1): 95-111.

489 490

Hellebust, J. A. (1965). "Excretion of some organic compounds by marine 491

phytoplankton." Limnol. Oceanogr. 10: 192-206.

492 493

Hellebust, J. A. (1974). "Extracellular products. In: W. D. P. Stewart, (Ed.), Algal 494

physiology and biochemistry." Blackwell: 838-863.

495 496

Hudson, R.J.M, et al. (1992). "Investigations of iron coordination and redox reactions 497

in seawater using ⁵⁹Fe radiometry and ion-pair solvent extraction of amphiphilic iron 498

complexes." Marine Chemistry 38: 209-235.

499 500

Hutchins, D. A. and K. W. Bruland (1998). "Iron-limited diatom growth and Si:N 501

uptake ratios in a coastal upwelling regime." Nature 393: 561-564.

502 503

Johnson, K. S., K. H. Coale, et al. (1994). "Iron photochemistry in seawater from the 504

equatorial Pacific." Mar. Chem. 46: 319-334.

505 506

Johnson, K. S., R. M. Gordon, et al. (1997). "What controls dissolved iron 507

concentrations in the world ocean?" Marine Chemistry 57(3 / 4): 137.

508 509

King, D. W., H. A. Lounsbury, et al. (1995). "Rates and mechanism of Fe(II) 510

oxidation at nanomolar total iron concentrations." Environ. Sci. Technol. 29: 818-824.

511 512

Kuma, K., S. Nakabayashi, et al. (1995). "Photoreduction of Fe(III) by 513

hydrocarboxylic acids in seawater." Water Research 29(6): 1559-1569.

514 515

Kuma, K., S. Nakabayashi, et al. (1992). "Photoreduction of Fe(III) by dissolved 516

organic substances and existence of Fe(II) in seawater during spring blooms." Mar.

517

Chem. 37: 15-27.

518 519

Laglera, L. M. and C. M. G. Van den Berg (2007). "Wavelength dependence of the 520

photochemical reduction of iron in arctic seawater." Environ. Sci. Technol. 41: 2296- 521

2302.

522 523

Macrellis, H. M., C. G. Trick, et al. (2001). "Collection and detection of natural iron- 524

binding ligands from seawater." Marine Chemistry 76: 175-187.

525 526

Miles, C. J. and P. L. Brezonik (1981). "Oxygen consumption in humic-colored 527

waters by a photochemical ferrous-ferric catalytic cycle." Environ. Sci. Technol.

528

15(9): 1089-1095.

529 530

Miller, W. L., D. W. King, et al. (1995). "Photochemical redox cycling of iron in 531

coastal seawater." Mar. Chem. 50: 63-77.

532 533

Millero, F. J. and S. Sotolongo (1989). "The oxidation of Fe(II) with H202 in 534

seawater." Geochim. Cosmochim. Acta 53: 1867-1873.

535 536

Millero, F. J., S. Sotolongo, et al. (1987). "The oxidation kinetics of Fe(II) in 537

seawater." Geochim. Cosmochim. Acta 51: 793-801.

538 539

Moffett, J. W. and O. C. Zafiriou (1990). "An investigation of hydrogen peroxide in 540

surface waters of Vineyard Sound with H218O2 and ¹⁸O2." Limnol. Oceanogr. 35:

541

1221-1229.

542 543

Morel, F. M. M. and N. M. Price (2003). "The biogeochemical cycles of trace metals 544

in the oceans." Science 300: 944-947.

545 546

Myklestad, S. M., E. Skanoy, et al. (1997). "A sensitive and rapid method for analysis 547

of dissolved mono- and polysaccharides in seawater." Marine Chemistry 56(3-4):

548

279-286.

549 550

Obernosterer, I. B. (2000). "Photochemical transformations of dissolved organic 551

matter and its subsequent utilization by marine bacterioplankton." PhD thesis: 133 pp.

552 553

Ozturk, M., P. L. Croot, et al. (2004). "Iron enrichment and photoreduction of iron 554

under UVand PAR in the presence of hydroxycarboxylic acid: implications for 555

phytoplankton growth in the Southern Ocean." Deep Sea Research II 51: 2841-2856.

556 557

Passow, U. (2002). "Transparent exopolymer particles (TEP) in aquatic 558

environments." Progress in Oceanography 55: 287-333.

559 560

Passow, U., J. Dunne, et al. (2006). "Organic carbon to ²³⁴Th ratios of marine organic 561

matter." Mar. Chem. 100: 323-336.

562 563

Petasne, R. G. and R. G. Zika (1997). "Hydrogen peroxide lifetimes in south Florida 564

coastal and offshore waters." Mar. Chem. 56: 215-225.

565 566

Price, D., R. F. C. Mantoura, et al. (1998). "Shipboard determination of hydrogen 567

peroxide in the western Mediterranean sea using flow injection with 568

chemiluminescence detection." Analytica Chimica Acta 377: 145-155.

569

570

Rice, P. J., B. E. Lockhart, et al. (2004). "Pharmacokinetics of fungal (1–3)-β-Image- 571

glucans following intravenous administration in rats." International 572

Immunopharmacology 4(9): 1209-1215.

573 574

Rijkenberg, M. J. A., A. C. Fischer, et al. (2005). "The influence of UV irradiation on 575

photoreduction of iron in th Southern Ocean." Mar. Chem. 93: 119-129.

576 577

Rijkenberg, M. J. A., L. J. A. Gerringa, et al. (2006). "Enhancement and inhibition of 578

iron photoreduction by individual ligands in open ocean seawater." Geochimica Et 579

Cosmochimica Acta 70(11): 2790-2805.

580 581

Rose, A. L. and D. Waite (2006). "Role of superoxide in the photochemical reduction 582

of iron in seawater." Geochimica Et Cosmochimica Acta 70(15): 3869-3882.

583 584

Rose, A. L. and T. D. Waite (2002). "Kinetic model for Fe(II) oxidation in seawater in 585

the absence and presence of natural organic matter." Environ. Sci. Technol. 36: 433- 586

444.

587 588

Rose, A. L. and T. D. Waite (2003a). "Effect of Dissolved Natural Organic Matter on 589

the Kinetics of Ferrous Iron Oxygenation in Seawater." Environ. Sci. Technol. 37:

590

4877-4886.

591 592

Rose, A. L. and T. D. Waite (2003b). "Kinetics of iron complexation by dissolved 593

natural organic matter in coastal waters." Marine Chemistry 84(1-2): 85-103.

594 595

Rose, A. L. and T. D. Waite (2005). "Reduction of organically complexed ferric iron 596

by superoxide in a simulated natural water." Environmental Science & Technology 597

39(8): 2645-2650.

598 599

Rue, E. L. and K. W. Bruland (1995). "Complexation of iron(III) by natural organic 600

ligands in the central North Pacific as determined by a new competitive ligand 601

equilibrium / adsorptive cathodic stripping voltammetric method." Marine Chemistry 602

50: 117-138.

603 604

Santana-Casiano, J., M. G.-D. Vila, et al. (2000). "The effect of organic compounds in 605

the oxidation kinetics of Fe(II)." Marine Chemistry 70(1-3): 211-222.

606 607

Santana-Casiano, J. M., M. Gonzalez-Davila, et al. (2004). "The oxidation of Fe(II) in 608

NaCl-HCO3- and seawater solutions in the presence of phthalate and salicylate ions: a 609

kinetic model." Mar. Chem. 85(1-2): 27-40.

610 611

Scully, N. M., D. J. McQueen, et al. (1996). "Hydrogen peroxide formation: The 612

interaction of ultraviolet radiation and dissolved organic carbon in lake waters along a 613

43-75 degrees N gradient." Limnol. Oceanogr. 41(3): 540-548.

614 615

Sedlak, D. L. and J. Hoigne (1993). "The role of copper and oxalate in the redox 616

cycling of iron in atmospheric waters." Atmospheric Environment 27A(14): 2173- 617

2185.

618 619

Sivia, D. S. (2006). "Data Analysis, A Bayesian Tutorial." 246 pp.

620 621

Stookey, L. L. (1970). "Ferrozine - a new spectrophotometric reagent for iron." Anal.

622

Chem. 42(7): 779-781.

623 624

Sunda, W. and S. Huntsman (2003). "Effect of pH, light, and temperature on Fe- 625

EDTA chelation and Fe hydrolysis in seawater." Mar. Chem. 84: 35-47.

626 627

Tanaka, Hurlburt, et al. (1971). "Application of Algal Polysaccharides as in vivo 628

Binders of Metal Pollutants." Proceedings of the International Seaweed Symposium 7:

629

602-604.

630 631

Theis, T. L. and P. C. Singer (1974). "Complexation of Iron(ll) by organic matter and 632

its effect on Iron(ll) oxygenation." Environ. Sci. Technol. 8: 569-573.

633 634

van den Berg, C. M. G. (1995). "Evidence for organic complexation of iron in 635

seawater." Marine Chemistry 50: 139-157.

636 637

Viollier, E., P. W. Inglett, et al. (2000). "The ferrozine method revisited: Fe(II)/Fe(III) 638

determination in natural waters." Applied Geochemistry 15(6): 785-790.

639 640

Voelker, B. M. and D. L. Sedlak (1995). "Iron reduction by photoproduced 641

superoxide in seawater." Mar. Chem. 50: 93-102.

642 643

Waite, T. D. and F. M. M. Morel (1984). "Photoreductive dissolution of colloidal iron 644

oxides in natural waters." Environmental Science & Technology 18: 860-868.

645 646

Waite, T. D., A. L. Rose, et al. (2006). "Superoxide-mediated reduction of ferric iron 647

in natural aquatic systems." Geochimica Et Cosmochimica Acta 70(18): A681-A681.

648 649

Wells and M. L. a. L.M.~Mayer (1991a). "The photoconversion of colloidal iron 650

oxyhydroxides in seawater." Deep-Sea-Research A 38: 1379-1395.

651 652

Wells, M., L. Mayer, et al. (1991b). "The photolysis of colloidal iron in the oceans."

653

Nature 252: 248-250.

654 655

Yocis, B. H., D. J. Kieber, et al. (2000). "Photochemical production of hydrogen 656

peroxide in Antarctic Waters." Deep Sea Research I 47(6): 1077-1099.

657 658

Yuan, J. and A. M. Shiller (1999). "Determination of subnanomolar levels of 659

hydrogen peroxide in seawater by reagent-injection chemiluminescence detection."

660

Analytical Chemistry 71(10): 1975-1980.

661 662

Yuan, J. and A. M. Shiller (2001). "The distribution of hydrogen peroxide in the 663

southern and central Atlantic ocean." Deep Sea Research II 48: 2947-2970.

664 665

Zepp, R. G. and D. M. Cline (1977). "Rates of direct photolysis in aquatic 666

environment." Environ. Sci. Technol. 11(4): 359-366.

667 668

7 Figures 669

time [min]

0 60 120 180 240 300

H2O2 [nM]

0 50 100 150 200 250 300

670

Figure 1: Photogeneration of H2O2 during 270 min of irradiation of a 10 mg L^-1 671

solution of laminarin (open triangle), carrageenan (open circle), gum xanthan (filled 672

circle) and of pure MQ water (filled triangle) and the mean of all 4 dark controls 673

(filled squares) 674

675

wavelength [nm]

100 200 300 400 500 600 700 800

normalised absorbance [abs L g-1 cm-1 ]

-0.2 0.0 0.2 0.4 0.6 0.8 1.0

676

Figure 2: Absorbance spectra (normalised absorbance for 1 g L^-1 and 5 cm cuvette) of 677

laminarin (dashed line), carrageenan (dotted line), gum xanthan (solid line) dissolved 678

in MQ water and filtered over 0.2 µm membrane 679

680

681

Figure 3: Dark oxidation of 218 nmol L^-1 Fe(II) in natural SW (filled circles) and 682

natural SW with PS added. Model results of oxidation of Fe (II) under O2 saturation 683

(dotted line) and in the presence of 5 nmol L^-1 H2O2 (solid line) at pH = 8.4, S = 34.1, 684

18 °C are also depicted 685

686

687

Figure 4: Dark oxidation of 230 nmol L^-1 Fe(II) in UVSW (filled circles) and UVSW 688

with PS added. Model results of oxidation of Fe (II) under O2 saturation (dotted line) 689

and in the presence of 270 nmol L^-1 H2O2 (solid line) at pH = 8.3, S = 34.1, 17 °C are 690

also depicted 691

692

693

Figure 5: Oxidation of Fe(II) in pure UVSW (triangles) and in UVSW with added 694

diatom exudates (circles) (22 °C, S = 34.2, O2 saturated, pH = 8.1, UVB = 0.3 W m^-2, 695

UVA = 17.6 W m^-2, PAR = 3.8 W m^-2). The dotted line depicts the detection limit.

696

697 698 699 700

701 702

Figure 6: Schematic photoredox cycle for FeL describing the Fe cycling in experiment 703

3 adapted from Sunda and Huntsman (2003) 704

705

0 500 1000 1500 2000 2500 3000 3500 4000

0 10 20 30 40 50 60 70 80 90 100

time [s]

Fe(II) [nmol L−1 ]

706

L

[Fe(II)]

[Fe(III)]

O

/[H

O

] L

khν

k^O2, k^H2O2

hν

[FeL]

Figure 7: Best curve fits for measured data (experiment 3) of the oxidation of Fe(II) in 707

UVSW (22 °C, pH = 8.1) with added diatom exudates (diamonds) using fix oxidation 708

rates calculated according to Millero et al. (1987; 1989) and the best estimate for the 709

photoproduction term (solid line) and using the best parameter estimates for all three 710

parameters (dashed line) the dotted line shows the detection limit 711